Thanks to the video for this talk, I've started writing my first translator, to help with Swedish newssites (since none of them are recognized as such). Hopefully first of many translators
@Czar Could you mind take a look at this code and help with with pointers? :)
I appreciate Sebastian's efforts, I really do. But this is precisely the type of project that the staff at the Wikimedia Foundation - or a contractor hired by the WMF - should be doing. It is grossly inefficient for a group of volunteers with limited time, and possibly programming skills that aren't quite aligned with what is needed, to each spend time figuring out how the Zotero citation system works, how to deal with various quirks (e.g., cookies), getting access to paywalled journals, and so on - grossly inefficient compared to one paid, experienced programmer working building Zotero translations for the top 100 or top 1000 sources.
If you want to see what happens when volunteers are relied on for this daunting task, see https://en.wikipedia.org/wiki/Wikipedia:TWL/Citoid#Results , which documents a set of unresolved problems, the first more than a year old.
Moreover, Zotero translators are used by far more than by Wikipedia - in fact, we're rather late to the game. So if the WMF hired someone to work on this issue, it would be contributing to the research community (and the greater good), not just Wikipedia.
I believe that you could look into finding a grant or another kind of scholarship for someone who could look into this kind of work. OTOH, there are volunteers who are able to work on this kind of stuff - User:Czar would be one, and I wonder if he has insights into the problems on the page you're linking to. (Finding this kind of gifted volunteers is precisely the scope of Help:VisualEditor/Community_Taskforce, whose development also foresees improving existing documentation so that it is at least clear what are the steps that anyone would need to do to start working in this area.)
I would recommend moving the aforementioned Results section to a centralized page with more traffic. I don't disagree that it would likely be more efficient to hire a programmer specifically for writing translators (especially if it looks like it will be a bottleneck in the Citoid development process) but I taught myself how to write Zotero translators because it's an open source project and the task is far bigger than one person. (Most of the issues on that page would not fall within a top 100 translators purview, possibly not even top 500.)
If there are any urgent cases on the TWL page, I'm happy to help out—just ping me. Otherwise, offhand, they don't look any more urgent than the backlog at https://github.com/zotero/translators/issues
